AI foundations concept

What is AI?

Reference

What is AI? Artificial intelligence is the field of building and studying computer systems that act in ways suited to a task and their surroundings. Some AI systems make predictions, recommendations, or decisions; others produce content. AI is a broad field, so the label alone does not say how a system works or how well it performs.

Reviewed 2026-10-09

AI means systems built to act intelligently

Artificial intelligence (AI) is the field that studies and builds computational agents that act intelligently. An agent is something that acts in an environment; a computational agent is one whose decisions can be explained through computation. To judge whether its behavior is intelligent, consider whether its actions fit its circumstances and goals, while allowing for its limits. In practice, AI systems can produce predictions, recommendations, or decisions that influence real or virtual environments. The definition describes a field and a range of systems, rather than one particular machine or method.

Some AI systems learn from data

There is no single recipe that explains how every AI system works. One common approach is machine learning (ML): a program trains a model from input data, and the trained model can make predictions on new data. Other AI systems can be described as computational agents: they take in information from an environment and use it when deciding what to do. In either case, the useful question is what the system is built to do and how it produces its output. A label such as “AI” by itself does not explain the system’s internal steps.

Machine learning is part of AI

AI is the wider field; machine learning is one subfield within it. The terms are sometimes used as if they mean the same thing, but that usage hides the relationship between them. When a system learns a model from data, it is using machine learning; the broader AI field also studies computational agents and their actions. Generative AI refers to systems that produce content, while predictive machine learning refers to systems not based on generative AI. For more on these approaches, see Predicting and generating in the AI foundations topic.

The label does not tell you everything

Calling something AI does not mean it works like a person or can handle every kind of task. AI systems differ in what they can observe, what actions they can take, and what goals they are designed to meet. Some systems are built for a specific task; the parent topic defines these as Narrow AI. It is more useful to ask what a particular system is meant to do, what information it can use, and how its performance is judged than to treat “AI” as a measure of general ability. The label describes a broad field, not a guarantee about any one system.

Questions people ask

What counts as AI?

AI includes computer systems designed to act intelligently or produce outputs such as predictions, recommendations, and decisions for a task.

Is AI the same as machine learning?

No. Machine learning is a subfield of AI. People sometimes use the terms interchangeably, but they describe different scopes.

Does AI always learn from examples?

No. Learning from data is one approach within AI. The broader field also studies computational agents and how they act.

Does an AI system have human-level abilities?

The label “AI” alone does not tell you how broad a system’s abilities are. Check the tasks it was designed for and how it performs on them.

Learn more

Artificial Intelligence: Foundations of Computational Agents
Its opening chapter introduces AI through computational agents, their actions, and their limits.
Machine Learning Glossary
Its definitions clarify how machine learning fits within AI and distinguish generative AI from predictive machine learning.

What changed

Sources

Part of AI foundations: What AI means Predicting and generating Capabilities depend on the task